en
0.25
0.5
0.75
1.25
1.5
1.75
2
Selective Sampling with Almost Optimal Guarantees for Learning to Rank from Pairwise Preferences
Published on 2012-01-242976 Views
Related categories
Presentation
Selective sampling with almost optimal guarantees for learning to rank form pairwise preferences00:00
Agenda00:11
To make the long story short00:26
Ranking from pairwise preferences :: Define (1/2)00:55
Ranking from pairwise preferences :: Define (2/2)02:00
Ranking from pairwise preferences :: Toy example03:19
(it’s all about the most relevant) Facts07:19
epsilon-smooth approximations :: Notation09:28
epsilon-smooth approximations :: Iterative algorithm10:55
epsilon-smooth approximations :: Selective sampling12:46
Selective sampling :: The eps-smooth estimator15:28
epsilon-smooth approximations :: Definitions17:10
epsilon-smooth approximations :: Algorithm analysis18:43
is eps-smooth19:32
SVM-Rank & Convex relaxations20:33
Conclusions & future work23:22